Skip to contents

Subgroup Treatment Effect Pattern (STEP) Fit for Survival Outcome

Source: R/fit_survival_step.R
fit_survival_step.Rd

[Stable]

This fits the Subgroup Treatment Effect Pattern models for a survival outcome. The treatment arm variable must have exactly 2 levels, where the first one is taken as reference and the estimated hazard ratios are for the comparison of the second level vs. the first one.

The model which is fit is:
Surv(time, event) ~ arm * poly(biomarker, degree) + covariates + strata(strata)
where degree is specified by control_step(). Note that for the default degree 0 the biomarker variable is not included in the model.

Usage 

fit_survival_step(
  variables,
  data,
  control = c(control_step(), control_coxph())
)

Arguments

variables

(named list of character)
list of analysis variables: needs time, event, arm, biomarker, and optional covariates and strata.

data

(data.frame)
the dataset containing the variables to summarize.

control

(named list)
combined control list from control_step() and control_coxph().

Value

a matrix of class step. The first part of the columns describe the subgroup intervals used for the biomarker variable, including where the center of the intervals are and their bounds. The second part of the columns contain the estimates for the treatment arm comparison.

See also

control_step() and control_coxph() for the available customization options.

Examples 


# Testing dataset with just two treatment arms.
library(scda)
library(dplyr)
library(rtables)

adtte <- synthetic_cdisc_data("latest")$adtte

adtte_f <- adtte %>%
  filter(
    PARAMCD == "OS",
    ARM %in% c("B: Placebo", "A: Drug X")
  ) %>%
  mutate(
    # Reorder levels of ARM to display reference arm before treatment arm.
    ARM = droplevels(forcats::fct_relevel(ARM, "B: Placebo")),
    is_event = CNSR == 0
  )
labels <- c("ARM" = "Treatment Arm", "is_event" = "Event Flag")
formatters::var_labels(adtte_f)[names(labels)] <- labels

variables <- list(
  arm = "ARM",
  biomarker = "BMRKR1",
  covariates = c("AGE", "BMRKR2"),
  event = "is_event",
  time = "AVAL"
)

# Fit default STEP models: Here a constant treatment effect is estimated in each subgroup.
step_matrix <- fit_survival_step(
  variables = variables,
  data = adtte_f
)
dim(step_matrix)
#> [1] 39 12
head(step_matrix)
#>      Percentile Center Percentile Lower Percentile Upper Interval Center
#> [1,]             0.025                0            0.275        1.240529
#> [2,]             0.050                0            0.300        1.628805
#> [3,]             0.075                0            0.325        1.818258
#> [4,]             0.100                0            0.350        2.003407
#> [5,]             0.125                0            0.375        2.300801
#> [6,]             0.150                0            0.400        2.638651
#>      Interval Lower Interval Upper   n events       loghr        se   ci_lower
#> [1,]      0.4055982       3.504396  74     41 0.006507558 0.3163349 -0.6134975
#> [2,]      0.4055982       3.686405  81     45 0.056543290 0.3031253 -0.5375715
#> [3,]      0.4055982       3.805456  87     50 0.118428018 0.2903633 -0.4506737
#> [4,]      0.4055982       3.925492  94     55 0.152132172 0.2808143 -0.3982538
#> [5,]      0.4055982       4.121412 101     59 0.099905333 0.2715645 -0.4323514
#> [6,]      0.4055982       4.284773 107     64 0.082851772 0.2574120 -0.4216664
#>       ci_upper
#> [1,] 0.6265126
#> [2,] 0.6506581
#> [3,] 0.6875297
#> [4,] 0.7025181
#> [5,] 0.6321620
#> [6,] 0.5873699

# Specify different polynomial degree for the biomarker interaction to use more flexible local
# models. Or specify different Cox regression options.
step_matrix2 <- fit_survival_step(
  variables = variables,
  data = adtte_f,
  control = c(control_coxph(conf_level = 0.9), control_step(degree = 2))
)

# Use a global model with cubic interaction and only 5 points.
step_matrix3 <- fit_survival_step(
  variables = variables,
  data = adtte_f,
  control = c(control_coxph(), control_step(bandwidth = NULL, degree = 3, num_points = 5L))
)